publisher colophon

This paper provides a philosophical analysis of self-knowledge or insight in psychotherapy. Our aim is to make sense of the idea—central to therapeutic practice—that insight can have a healing effect. Specifically, we focus on the following question: How should a person relate to his or her own (maladaptive) mental states, such that the resulting self-knowledge constitutes therapeutic self-knowledge, that is, self-knowledge that facilitates therapeutic change? We propose a ‘dual perspective’ account of such therapeutic self-knowledge. According to this account, genuine insight into one’s (dysfunctional) beliefs, desires, or emotions requires that one avows these states from a first-person or point of view, while at the same time regulating this avowal from a more adaptive, second-person perspective on oneself that mirrors the therapeutic relation.


Self-knowledge, insight, psychotherapy, first-person authority, second-person perspective, self-regulation, Moran

The development of self-knowledge or self-insight is a well-recognized therapeutic factor in psychotherapy. In some way or other, all evidence-based therapies seek to reframe and enrich patients’ own understanding of themselves. In this article, we focus on self-knowledge with respect to mental states (beliefs, emotions, etc.), in particular those (maladaptive) states that cause patients to seek treatment.

As an example, imagine a person who enrolls in psychotherapy because he finds himself unable to commit himself to intimate relationships. During the first session, he tells his therapist about his youth, about how his father abandoned him and his mother when he was 8 years old. At one point he says, “I guess I’m afraid of being abandoned again, that’s probably why I panic every time someone gets too close.” Therapy begins and they make a good start. A constructive therapeutic alliance is formed, and the patient unconsciously starts to become attached to his therapist. As therapy progresses, however, he starts to feel more and more anxious during and in-between sessions, up to the point where he cancels his session for the first time. During the following session, they discuss what happened and he comes to understand that the anxiety he presently feels is the result of having become attached to his therapist and signifies his fear of being abandoned by him when therapy ends. He realizes that his behavior of canceling their previous session fits right into the pattern that troubled him before and made him seek treatment in the first place. At the end of the [End Page 45] session he says, “I already knew I had abandonment issues—I remember me telling you about it during our first session—yet up till now, I somehow never really understood, I never really knew.”

In the course of therapy, patients typically come to understand some of their mental states ‘from a new perspective.’ But what does this mean, exactly? And how should we understand the idea that relating to one’s own mental states from such a new perspective can lead to therapeutic change? Self-knowledge regarding one’s mental states comes in many forms, not all of which are therapeutically helpful. One can know about one’s mental states in a more or less ‘theoretical’ way, for example, through reading about it in a psychology book or listening to the folk theories and advice of others, and on that basis make a conjecture about one’s own state of mind. In this sense, our fictive patient ‘already knew’ about his fear of abandonment.

Whereas such ‘third-person’ self-knowledge may motivate one to seek treatment, it will not by itself constitute the kind of self-understanding that therapists generally want to achieve with their patients. The alternative is to know one’s mental states experientially, namely, knowing one’s fears in this way involves being fully absorbed by them in moments of panic or distress. But without further qualification, this alternative ‘first-person’ conception does not explain how self-knowledge can constitute a therapeutic factor. Gaining knowledge of one’s own fears, doubts, painful memories, or paranoid beliefs from a first-person point of view can be quite harmful. It may lead to feelings of helplessness or despair, for example, making one quit therapy, or worse.

Our aim in this article is to make sense of the idea—central to therapeutic practice—that insight can have a healing effect. Specifically, we focus on the following question: How should a person relate to his or her own (maladaptive) mental states such that the resulting self-knowledge constitutes therapeutic self-knowledge, that is, self-knowledge that facilitates therapeutic change? We are, thus, primarily interested in the philosophical question of how to conceive of insight such that it implies having therapeutic potential. We will not target specific empirical questions, such as what patients need to learn about themselves, how self-knowledge needs to be brought about through specific therapeutic interventions, or by means of which psychological mechanisms it leads to improvement of symptoms or behavioral change. In developing our account, we, therefore, primarily engage with philosophical accounts on self-knowledge and apply them to psychotherapeutic practice.

The article is structured as follows. In the next section, we introduce our target phenomenon by outlining a central working principle in psychotherapy: the principle of first-person authority regarding the patient’s own mental states. Our question is how to understand therapeutic self-knowledge in such a way as to honor this principle. In the next section, we discuss Richard Moran’s (2001) well-known ‘agency account’ of self-knowledge. On this account, first-person authority is to be understood in terms of a person’s capacity to actively engage with and shape her own mental states. Moran explains the first-personal authority we have regarding our own mental states exclusively in terms of 1) transparent avowal of one’s mental states by means of 2) rational deliberation. We take issue with both claims. In the subsequent section, we argue that, owing to his exclusive focus on reason, Moran cannot properly account for authoritative self-knowledge of feelings and emotions. Our main concern, however, lies with the centrality of transparency in Moran’s account. Building on critical papers by Lear (2004) and McGeer (2008), we then go on to show how transparent avowal of one’s mental states can severely undermine one’s self-understanding and, when going unchecked, can pose a real obstacle in treatment. We argue that one also needs to be able to regulate one’s transparent avowals from an allocentric, second-person perspective on one’s own mental states. In the penultimate section we, therefore, suggest that therapeutic self-knowledge should be characterized in terms of adopting a dual perspective one’s mental states: the state of ‘really knowing’ one’s mind, as expressed by our patient as described, involves both the transparent avowal of the relevant mental states as well as the acknowledgment and regulation and of this avowal from a second-person perspective on oneself. In the final section, we briefly mention [End Page 46] some interesting touching points with conceptions from the psychotherapeutic literature and end with some directions for future research.

The Principle of First-Person Authority

One of the goals of psychotherapy, no matter which specific orientation, is to increase patients’ understanding of their own mental states, in particular those states that lie at the basis of the maladaptive emotional and behavioral responses that make them seek treatment. What do I really feel when I respond dismissively to others? What does my emotional reaction tell me about what I believe about myself? Should I believe the thoughts that run through my head when I start feeling anxious? Why do I always try to please others? What am I trying to avoid by behaving this way? And so on. A significant part of psychotherapy consists in collaboratively finding the right answers to such questions. This process of collaborative exploration, discovery, and interpretation of the patients’ mental states is guided by a basic working principle. All therapists, no matter their specific orientation, seem to honor it.

We term this principle the principle of first-person authority regarding the patient’s mental states. The patient, not the therapist, is considered to be in a privileged position to speak her mind during therapy sessions. As a practical rule, the therapist does not question or criticize the patient’s answers when asking her what she feels, thinks, wants, or believes. Even if the therapist is under the impression that the patient is still unaware of a certain painful emotion, or mistakes one type of mental state for another (e.g., disappointment for anger), the patient’s self-reports are normally not discarded, but rather taken as further stepping stones toward improving her self-understanding.

The principle of first-person authority is reflected in the open-minded, encouraging, inquisitive, and curious attitude therapists adopt toward their patients’ stories and ideas about themselves, others, and the world. In the psychotherapeutic literature, it is reflected for example in the notion of the therapist’s ‘not-knowing stance’ toward the patient’s mental life (e.g., Fonagy & Bateman, 2012). Of course, patients typically lack self-knowledge regarding some of their core mental states. This is often part of the reason why they seek treatment. Experienced therapists quickly form hypotheses about what is going on in their patient’s mind long before the patient comes to know and understand this. What the principle of first-person authority says, however, is that, ultimately, the validity (therapeutic significance) of such hypotheses is conditional on the patient’s endorsement.

What does this principle tell us about the nature of the particular kind of self-knowledge that psychotherapy aims for? It is a datum in psychotherapy that the patient needs to be actively involved in the development of her own self-insight for it to become therapeutically useful. Simply telling the patient, as a passive recipient, about her maladaptive thoughts and feelings generally does not work. The trick is to guide the patient toward her own discovery of her problems (as reflected, e.g., in the notion of ‘Socratic questioning’ in cognitive therapy [Padesky, 1993]). Therapeutic self-knowledge develops through the active engagement of the patient in coming to know her own mind. It requires a process of appropriation of some kind, in which the patient makes her own evaluation about what she thinks and feels. This much seems to be required for the patient to start acknowledging some maladaptive mental state ‘as her own,’ in the sense that it is hers to come to terms with and hers to manage in ways she considers appropriate. But how should we understand this?

Moran on Authoritative Self-Knowledge

Central to Moran’s (2001) philosophical account of self-knowledge is the idea that authoritatively knowing our own mental states, for example, finding out what we believe or want, is a matter of actively making up our minds—not in an arbitrary way, but rather by deliberating on their particular subject matter.

Moran sets out to explain the special nature of first-person authority in the wider context of (moral) human agency. He criticizes accounts of [End Page 47] self-knowledge that seek to explain first-person authority exclusively in epistemic terms (2001, p. 1). Such ‘epistemic’ accounts typically model first-person knowledge on perception or theory-based observation (e.g., Armstrong, 1981; Dennett, 1987; Gopnik, 1993; Carruthers, 2011). In cases of perception or theory-based observation, the perceiver or observer typically has a passive role in relation to the object perceived or observed, in the sense that the object leads an existence independent of the epistemic subject. Moran, by contrast, argues that such passivity is alien to our conception of ourselves as self-knowing agents. To use one of Moran’s examples (2001, p. 26), imagine asking someone whether she intends to pay back the money she borrowed. Suppose she answers, “As far as I can tell, yes.” What makes this response particularly disturbing is that it seems to be issued from an onlooker’s perspective, as if she were talking about someone else. We generally do not accept such answers precisely because they signal a lack of first-person involvement. We demand that others play an active part in coming to know their own mental states; we demand that they make up their mind, that is, decide whether they shall pay back the money, perhaps after deliberating about the matter and endorsing the intention to (not) do so.

Moran, thus, criticizes purely ‘epistemic’ accounts of self-knowledge for modeling the self-relation as an exclusively third-person or ‘theoretical’ stance toward one’s own mental states. Thus, he states:

What is left out of the Spectator’s view is the fact that I not only have a special access to someone’s mental life, but that it is mine, expressive of my relation to the world, subject to my evaluation, correction, doubts and tensions.

(p. 37, emphasis in original)

So what, then, are the conditions for agential self-knowledge of the kind that Moran proposes? The answer to this question, for Moran, lies in the transparency of rational deliberation. According to what he calls the ‘transparency condition’ of first-person statements of, for example, one’s beliefs, one should treat the question of one’s belief about P as equivalent to the question of the truth of P (pp. 62–63). Drawing from Edgly (1969) and Evans (1982), Moran argues that this equivalence should not be understood in a strictly logical sense; the question whether one believes that P does not somehow reduce to the question whether P. Rather, it means that one defers answering the self-directed question, “Do I believe that P?” to answering the world-directed question, “Is it the case that P?” This second, world-directed question (“Is it the case that P?”) is a deliberative question, which is issued from the first person. Answering such deliberative questions requires that one take an active stance toward the subject matter (P), by thinking about it and judging whether or not it is the case. Moreover, such world-directed questions are paradigmatically first personal: Others cannot settle our beliefs in the same way that we can, when we deliberate about some subject matter (whether it is raining, whether to go to Italy on holiday, whether to quit one’s job). Answering a deliberative question can, thus, only have a constitutive effect on the relevant mental states in the first-person case.

On Moran’s account, deferral to deliberation about the subject matter of our beliefs explains first-person authority. For insofar as we are rational agents, considering our reasons for or against believing P upon deliberation, actually determines whether or not we believe that P.1

Moran’s understanding of authoritative self-knowledge in terms of deliberative avowal is a good starting point for thinking about self-knowledge in psychotherapy.

First, Moran’s critical remarks regarding the incompleteness of purely epistemic accounts of self-knowledge make much sense from a therapeutic perspective. It is a common experience for therapists to hear patients talking about their own mental states in a rather detached way, ‘as if they were talking about someone else.’ Although epistemically, their knowledge claims may be perfectly adequate, such a disengaged stance toward oneself is generally considered therapeutically ineffective.

Second, Moran has a relatively simple answer to the question why the principle of first-person authority is at work in psychotherapy. In the previous section, we suggested that this principle reflects the fact that, for a particular judgment about the patient’s mental states to constitute self-knowledge [End Page 48] with therapeutic potential, it somehow needs to pass the verdict of the patient herself. On Moran’s account, this can be further unpacked as the patient’s deliberative avowal of the relevant mental states. We can then understand the principle of first-person authority with reference to the idea that it is precisely to the extent that the patient is ready to adopt this active attitude of avowal toward her own mental states that she can be said to have first-person authority with respect to those states.

The notion of avowal also gives us a first rudimentary understanding of the therapeutic potential of self-knowledge: Deliberative avowal implies a normative stance of commitment toward the relevant states. Consider a depressed patient who, with the help of her therapist, becomes consciously aware of a previously implicit core belief, for example, that she is worthless as a person. Upon reflection, she comes to the conclusion this belief is false and that, in fact, she is basically OK as a person. Authorizing this contrary belief through deliberative avowal can be regarded as a turning point in therapy in the sense that it marks the patient’s commitment to a new (and more adaptive) appraisal of herself. She has put herself in a position from which her maladaptive ways of thinking about herself can start to seem unauthorized, because they are in conflict with her newly avowed belief. And this, in turn, opens the door toward changing them (cf. Bilgrami, 2006). Moran explains first-person authority by avowal exclusively in terms of rational deliberation, however. As a result, avowal of non–belief-like states, in particular feelings and emotions, seem to fall outside the scope of Moran’s account. Yet, coming to terms with one’s feelings and emotions is crucial in psychotherapy and everyday life. In the next section, therefore, we focus on other, nondeliberative ways to make sense of the notion of authoritative avowal.

Nondeliberative Avowal of Feelings and Emotions

Moran suggests we treat self-knowledge of other kinds of attitudes in the same way as we treat beliefs, namely, by reflection on reasons and to make a judgment as to what we are to believe, in light of those reasons. Indeed, sometimes the deliberative stance seems to apply to our affective states, for example, when one comes to see one’s feelings of guilt as inappropriate or one’s anger as childish. On Moran’s view, judgments of this kind help to redetermine or shape the emotion, because “the ‘seeing’ in question is … an expression of the ordinary deliberative reflection about how to feel” (2001, p. 59, emphasis in original).

In many instances, however, our emotions turn out to be quite unresponsive to reasons. Consider the case of phobias. Moran brings it up as an example of what he too considers a familiar fact about some emotional states: that they are irrational and “do not alter when the beliefs on which they are based are sincerely denied by the person in question” (2001, p. 54). However, does this imply that the person in question does not have first-person authority with respect to his fear? Moran seems to be committed to this idea when he claims that the authority of self-knowledge “will only be in place to the extent that the person’s reasons really do determine what his beliefs and other attitudes are” (2004, p. 457; emphasis added).

But this position seems to be untenable. Of course, people sometimes do misidentify their own feelings or emotions, for example, mistaking a burst of panic for anger. And, sometimes, others are in a better position to know what we are feeling. But what we are considering here is the proposition that one lacks first-person authority regarding one’s emotions because and insofar as these emotions are irrational, unresponsive to reason. It is one thing to say to a person with acrophobia that he should not be afraid of heights because there is no reason to be afraid. But now imagine telling him that he is in no privileged position to know whether he is afraid of heights (and/or afraid of heights), for the reason that his attitudes toward heights are irrational. It goes against the basic normative structure of our (moral) social practice to deny a person first-person authority over his mental states solely on the grounds that they turn out to be unresponsive to reason. Consider cases like, “I know you did not mean it, but I am still angry with you,” or, “You are right, I should not care so much, but I still feel sad for him.” In general, [End Page 49] we would not withhold first-person authority regarding such irrational anger or sadness.

First-person authority regarding irrational feelings and emotions lies at the basis of psychotherapy. To identify core maladaptive (and often irrational) states that need to be targeted in therapy, therapists heavily rely on the patients’ own reports on what they think and feel. At this stage in therapy, whether these mental states are rational or not, is beside the point. To collaboratively establish that this is in fact the (irrational) feeling or emotion that is troubling the patient, therapists heavily rely on what we termed the working principle of first-person authority.

The reason why Moran’s treatment of deliberative avowal works so well for belief is that, insofar as we are rational, deliberation determines what we believe. Assuming rationality, this constitutive aspect can explain why we treat ourselves and each other as having authoritative self-knowledge of some of our beliefs. The problem is that attitudes are often not rational and fail to align with reason, even after considerable effort. As long as the constitutive role of agential self-directedness is understood exclusively in terms of deliberative avowal, Moran’s agency account cannot explain why knowledge of many of our affective and conative attitudes can have genuine first-person authority.

So, might there be other ways to characterize this agential aspect of self-knowledge for nondoxastic states? Charles Taylor (1985) has offered a helpful model in terms of which we know our own states through self-interpretation, that is, via a method that could, but need not be, deliberative in nature. Taylor characterizes the process of self-interpretation as acts of ‘re-describing,’ ‘finding new formulations,’ or ‘seeing in a different light’ (Taylor, 1985). He argues that, as ‘self-interpreting animals,’ we are bound up in a process of articulating ourselves, and that such self-interpretations shape us by informing and changing our attitudes. An attitude of resentment, for example, changes when one comes to see it as unjust. Moran discusses Taylor’s account, but criticizes it precisely for lacking the ingredients to explain how the interpretation of our own attitudes can truly be said to (re-)constitute those attitudes. Although Moran seems to be attracted to the idea of constituting our states through an act of self-interpretation, he thinks that Taylor’s characterization makes self-interpretation seem too arbitrary to do real constitutive work.

The critique, at the bottom line, is that one can redescribe the way one feels without in any sense feeling committed to this new description. In other words, our acts of self-interpretation might just be labels we can use to describe some of our emotions, but that do not have the power for us to be committed to the fact that this is, in fact, the state we are in.

Moran has a point here: Merely relabeling an emotion does not alter it. One has to believe that the new formulation is somehow more accurate than the old one, closer to the truth. But Taylor does not deny this. On his account, we first give the new description a ‘try’ as it were, to see whether or not our feelings and emotions become more focused, intelligible, or coherent to us as a result (or the opposite). The point is that it is not within our conscious control what we feel in response to a particular articulation of an emotion. Nor can we fully control when and under what circumstances our emotions resonate with our articulations. The question of whether and when a particular articulation actually makes the emotion seem to be more focused and determinate to us cannot be decided by a mere act of will. There are ‘empirical’ or ‘natural’ constraints on articulating our feelings that make the process of self-interpretation far from arbitrary. Accordingly, if we find that a new description resonates with our emotions and makes them seem more clearly circumscribed and determinate, this will give us reason to believe that this new description is more accurate than the old one. On Taylor’s account, then, redescribing one’s anger as disappointment or as jealousy, for example, will only have a transforming impact if one concludes that the proposition, “I am in fact disappointed/jealous” is closer to the truth than “I am angry.” This supposition is in line with Moran’s statement that, “[f]or a new description of my emotion or belief is powerless to alter it unless I believe the description. Clearly, it is the actual believing that is crucial to this change, and not the activity of naming or describing” (2001, p. 55). [End Page 50]

There is also a stronger reading of this passage, however, that goes beyond Taylor’s account of self-interpretation. On this reading, one not merely has to believe that one’s redescription of the emotional state one is in (e.g., jealousy instead of anger) is true, one also needs to rationally endorse its content. Accordingly, it is not enough that you come to believe that you actually experience a faint feeling of fear (rather than embarrassment) in the vicinity of a colleague, you must also believe that your colleague is in fact to be feared for this new description to authoritatively change the way you feel. Moran seems attracted to this stronger claim when he writes, “When the articulation or interpretation of one’s emotional state plays a role in the actual formation of that state, this will be because the interpretation is part of a deliberative inquiry about how to feel, how to respond” (2001, pp. 58–59, italics added).

We think this second reading is implausible. It is a common experience (especially in psychotherapy) that one rediscovers one’s feelings and emotions without (and perhaps precisely because one is not) feeling committed to the truth of their content upon reflection. As Carmen (2003, p. 404) observes, often the point of self-expression is not “to decide how I ought to feel, but to get clear about how I do feel by letting my emotions take shape and find a voice in what I say and do.” In psychotherapy, focusing on the quality of one’s emotion rather than on the truth of its content often has a liberating effect; it enables patients to focus on themselves and their emotions rather than become absorbed by the distorted picture of the world as represented by these emotions.

Following Taylor, we conclude that the self-constitutive effect of self-interpretation of our mental states need not require the rational endorsement of their content. It is enough that we come to experience the accuracy of our new descriptions on the basis of what these words stir up in us. This, we suggest, is what often happens in psychotherapy, when patients reflect on themselves in an attentive, receptive, inquisitive, perhaps creative and associative, yet not necessarily critical, frame of mind.2 This process of self-interpretation can also be augmented by nonlinguistic forms of self-expression, for example, through bodily gestures as well as music, art, dance, or sports. In mental health care, therapeutic formats such as creative therapy, music therapy, and psychomotor therapy have been explicitly designed for this purpose.

On Taylor’s account, one answers a question as to how one feels about some proposition (or something or someone) by actively engaging oneself with it and, through one’s (linguistic) expressions, steering oneself into a position in which one can feel it in a more determinate and articulate form. The result is a kind of avowal of expressed attitude that is affective in nature, not deliberative (e.g., “Yes, what I am currently feeling toward him is jealousy”). Such affective avowals, we argue, can play an equally important constitutive role in determining the attitude that is being avowed. We, therefore, see no reason why affective avowals should not be granted the same degree of first-person authority as their deliberative counterparts.

Allocentric Regulation of Transparent Avowals

There is another, more serious concern about Moran’s exclusive focus on transparent deliberation in accounting for first-person authority. This worry does not concern the deliberative aspect of Moran’s transparency condition on authoritative self-knowledge, but rather the requirement of transparency itself. The problem is that, when trying to find out what we feel, want, or believe, we cannot always trust our transparent outlook on the world. Jonathan Lear (2004) provides an illuminating discussion of ‘Mr. A,’ a borderline patient, who constantly interprets events in his life under the concept of betrayal. For instance, Mr. A will interpret his girlfriend’s angry outburst as a long-expected exposure of her true feelings toward him, and criticizes himself for having made himself so vulnerable for this betrayal in the first place (2004, p. 450). The problem for Mr. A, as Lear explains, is that the reasons he gives to support his beliefs and feelings fulfil a constraining function, rather than one that facilitates (rational) freedom. The more Mr. A transparently avows, the more locked in he will become in his interpretative schemas: [End Page 51]

For Mr. A, self-conscious reflection is a manifestation of his unfreedom: for as he reflects on his reasons for feeling betrayed, he digs himself ever deeper into a crabbed and constraining world. Rational, self-conscious reflection is being deployed as a defense, one which helps sustain the betrayed world. For him, avowal is the culmination of his imprisonment.

(Lear, 2004, p. 453)

Here we have a clear illustration of how, in psychoanalytic terms, deliberative avowals can function as a defense mechanism that effectively blocks more appropriate forms of self-understanding. But the case of Mr. A also points to the potentially undermining effect of affective avowals. In situations in which Mr. A feels most betrayed, affective avowal of his emotions will fully consume him, making him blind to anything else that is going on inside and around him, sometimes with dire behavioral consequences (e.g., aggression, self-harming behavior, suicide attempts).

The case of Mr. A indicates how avowal as such may have an undermining effect on self-knowledge and obstruct therapeutic progress. To appreciate this point, it is important that we distinguish between two aspects of first-person authority: the security of a self-ascription of a particular mental state through transparent avowal and the adequacy of this self-ascription when placing the mental state in the diachronic context of the self-ascriber’s mind. On the one hand, Mr. A can be said to know perfectly well what he currently feels or believes through transparent avowal, that is, that he feels betrayed. His therapist would not argue with Mr. A that this is what he feels, and in this sense, Mr. A’s avowal enjoys an important degree of security. Yet, on the other hand, his avowal lacks what we call ‘adequacy’ in the sense that it is not an adequate representation of his psychological situation and, therefore, lacks the authority to speak for him as a person. The avowal of betrayal effectively blocks, at least at that moment, the assessment and avowal of other (and more adaptive) feelings, desires, and beliefs that he also (and perhaps also currently) has, but that contrast with the general theme of betrayal, for example, his love for his girlfriend, his desire for personal intimacy, his belief that most people generally mean well, and so on. Mr. A’s paranoid avowal lures him into treating his feeling of betrayal as representative of how he feels in general and to act correspondingly. His paranoid avowal is a form of self-deception. Although he does have the mental states he avows (and in this sense has ‘secure’ self-knowledge of these states), he overestimates their value and importance relative to the full mental context in which they operate (and in this sense lacks ‘adequate’ self-knowledge of them).3

The (in)adequacy of transparent avowals is a major focus in therapy. In the treatment of borderline personality disorder, for example, treatment mainly consists in teaching patients, not how to (dis)avow their mental states, but rather how to regulate these attitudes and their inclinations to avow them in situations of arousal or crisis. There are well-established emotion regulation training programs in which patients learn to identify their emotions by level of intensity and to implement a co-authored personalized crisis management plan. This plan specifies for each level of emotional intensity what to do to decrease emotional distress.4 In psychotherapeutic treatments for borderline personality disorder, patients learn how to distinguish between different ways they are disposed to think and feel about themselves and others and to identify these as driving their self-experience and interactions with others. The goal is not primarily to avow or dismiss any such pattern as it threatens to unfold in real life, but to learn to regard them as such, that is, as patterns in one’s attitudes, allowing oneself to take a step back, to adopt the perspective of an attentive, respectful observer, and from that perspective to try to manage one’s attitudes in more adaptive ways.5

In this context, Victoria McGeer’s (e.g., 1996, 2008) agency account of self-knowledge is particularly instructive. McGeer agrees with Moran that deliberative avowal is an important road toward achieving first-person authority regarding our mental states. But she argues that, in many instances, it is not sufficient to authorize our self-ascriptions. Our deliberative powers are also known to be fragile, all too easily hijacked by our more subversive inclinations. Hence, she argues, our folk-psychological practice is designed not only to facilitate people to make reason-based [End Page 52] avowals, but also to teach them how to regulate their inclinations to avow.

In most cases, our practices of self-regulation are relatively effortless and automatic. There may be circumstances, however, in which “our self-managing techniques need to be explicit, effortful and well-planned” (McGeer, 2008, p. 89), for example, how to steer oneself into or out of a position in which avowal comes easiest, amplifying one’s deliberative capacities in moments of boredom or distraction, or rather sidetracking them when under great stress. Self-regulation in this sense is, thus, not a special means through which to avow the contents of one’s transparent world-directed attitudes, but rather a way of relating to such transparent avowals.

Importantly, self-reflection of this kind does not involve reflection on one’s avowals by having yet another go at deliberating about their subject matter, weighing reasons pro and contra. Rather, it draws attention to one’s dispositions to avow under certain circumstances. To see this, notice that, for example, understanding one’s own belief as a belief, that is, as a representational state, requires that one distinguish between the state of affairs represented and one’s (mis)representation of it. One needs to understand the content of one’s belief as a representation of something that possibly lies beyond one’s transparent point of view (cf. Vierkant, 2012). Relating to (patterns in) one’s attitudes, not merely their content, requires that one adopt an allocentric, second- or third-person perspective toward oneself. Crucially, the relevant (first-order) mental states are not transparent from the allocentric perspective; they remain opaque.6

We conclude that, given the fact that human beings are susceptible to ‘pathologies of avowal’ (Lear, 2004), both of the deliberative and affective kinds, allocentric self-regulation as a way of reflecting on and managing one’s dispositions to avow turns out to be an empirically necessary requirement for achieving the kind of authoritative self-knowledge that supports therapeutic progress.

The strengths of adopting an allocentric self-perspective—seeing oneself and one’s thoughts and feelings from a certain sort of distance—is, however, also its weakness, because it creates a risk of allowing for too much distance. We agree with Moran that an exclusively allocentric self-relation leads to self-alienation. In clinical practice, extreme cases of such self-alienation can be found in patients with symptoms of dissociation. Such patients may, for example, experience themselves as if not being ‘inside’ their own body, as if looking onto themselves from another point of view. These patients often report not being able to feel what is going on inside their body. As a consequence, they may have major difficulties in experiencing certain emotions, such as anger. Here, a self-alienating allocentric self-relation blocks (adaptive) affective avowal of (first-order) feelings and emotions. A less dramatic form of dissociation can be seen in patients (e.g., with borderline personality disorder) who ramble on about their own mental states in ways that seem totally disconnected with the actual experience and avowal of these states (see also final section). In these cases, what is missing is an act of first-personal (deliberative or affective) avowal of the relevant first-order mental states. We, thus, agree with Moran insofar as such avowal from the first-person perspective is necessary for patients to recognize and experience their maladaptive mental states as their own, in the strong sense of it being a result of their own self-engagement that the contents of these states can be felt or judged appropriately.

In psychotherapy, neither of these perspectives leads to a therapeutic self-relation all by itself. An exclusively allocentric self-perspective is self-alienating, whereas an exclusively first-person self-perspective is self-deceiving. For therapeutic self-knowledge, it is essential to find the right balance between the capacity to avow from a first-person perspective, and the capacity to address one’s avowals from an allocentric perspective. What is required, we suggest, is a dual perspective on self.

Therapeutic Self-Knowledge from a Dual Perspective

Let us briefly sum up our argument thus far. We started from the observation that the therapeutic potential of self-knowledge regarding the patient’s own mental states is intimately connected with the working principle of first-person authority: An insight only seems to facilitate therapeutic [End Page 53] change insofar as the patient herself authorizes it. We discussed Moran’s account of self-knowledge and explained why it is a good starting point for understanding the connection between the principle of first-person authority and the therapeutic potential of self-knowledge. First-person authority, understood in terms of deliberative avowal of one’s mental states from the first-person point of view, comes with a commitment to change one’s ways and bring one’s dispositions into accord with the states avowed.

Moran’s account, however, does not have the resources to explain first-person authority with regard to irrational mental states, as we argued subsequently. This is a significant omission, given the importance of patients’ first-person authority in closing in on therapeutically significant irrational thoughts, feelings, and emotions in collaboration with their therapist. We proposed Taylor’s account of self-interpretation as an alternative and introduced the notion of ‘affective’ avowal of such irrational states. But avowal, whether deliberative or affective, is not enough to account for the authority of therapeutic insight. When going unchecked, avowals, both of the deliberative and affective kinds, can severely undermine patients’ first-person authority and obstruct therapeutic progress. As we argued in the last section, a significant portion of psychotherapy consists in addressing ‘pathologies of avowal’ from an allocentric perspective on oneself.

How, then, should we understand the patient’s authoritative role in achieving self-knowledge with therapeutic potential? In this section, we suggest that therapeutic self-knowledge requires a dual perspective on oneself, consisting in 1) avowal of the relevant first-order states from the first-person point of view, and (2) acknowledgment and regulation of this avowal from a generally benevolent, second-person perspective on these first-order states.

Consider our patient from the introduction again, the man who seeks treatment because he cannot commit himself to serious relationships and who experiences increasing levels of anxiety as he starts to become attached to his therapist. Merely affectively avowing his fear to be abandoned by his therapist (or deliberatively avowing his reasons for fearing abandonment) from the first-person point of view will not automatically result in the breakthrough he experienced during the next session. For this avowal could just as well lead him to stop therapy or become angry with his therapist for abandoning him (analogous to Mr. A). Rather, what happens when our patient says that he now ‘really knows’ that he has abandonment issues, we suggest, is that his avowal of his present fear of being abandoned by his therapist is accompanied by an allocentric acknowledgment of the fact that the fear he feels in this particular instance is part of a larger destructive pattern that prevents him from achieving the level of intimacy he longs for in serious relationships. What makes his experience during this session genuinely revealing to him is that, for the first time in his life, he can now fully experience and avow his fear from the first-person point of view (‘I feel terrified that he will leave me behind!’) while also acknowledging its meaning and role in the larger context of his life from an allocentric perspective (‘This [kind of feeling] is what has been holding me back’). It is through this bi-perspectival appropriation of his fear, we propose, that our patient’s insight can be understood as a therapeutic factor. He has reached a position from which he can address, from an allocentric perspective, the fear he now experiences in its full complexity through avowal from the first-person perspective.

As we have seen, Moran dismisses accounts of self-knowledge that portray our relation to our own mental states as analogous to a perceiver and the object perceived or a scientist and the subject of inquiry, that is, as a purely empirical relation. We do not think all varieties of the allocentric self-relation come down to such an ‘empirical’ self-relation, however. The analogy with perception or scientific inquiry is misleading. We regard self-regulation primarily as a second-person relation, modeled on the interaction between people. Accordingly, allocentric self-regulation generally involves a kind of normative and affective engagement with oneself that is similar to that which one can have with other people.7 In therapy, the aim is to realize a shift toward second-personal self-directed attitudes such as love, concern, and compassion (and away from attitudes like disgust, hatred, and contempt). [End Page 54]

This proposal fits well with theorizing in psychotherapy. Several psychotherapeutic frameworks make reference to a process of ‘internalization’ to explain patients’ (mal)adaptive thoughts and feelings about themselves and their own mental states. The general idea is that people incorporate the interaction patterns they experienced with significant others in childhood (e.g., parents) to the effect that they come to experience themselves in ways that fit these patterns in adulthood.

Consider a father who responds to his 7-year-old son’s shame, disappointment, and frustration after wetting his bed once again, with a mixed attitude of love, acceptance, trust, and encouragement. The father takes an affectively laden, normative stance toward his son’s feelings from which he (implicitly or explicitly) conveys, for example, that it is okay for his son to feel this way, that there is no need to be afraid of these feelings, that it need not determine how he feels about himself, need not undermine his resolve, and so on. The boy can learn (through such interactions) to adopt a similar attitude toward his own shame, disappointment or frustration in later life, regulating these feelings by endorsing similar comments (that it is okay to feel this way, etc.) from an ‘internalized,’ second-person perspective on himself.

This is also how we would understand our fictive patient’s allocentric acknowledgement of his affectively avowed fear: ‘This is what has been holding me back.’ There seems to be an aspect of deliberation present here, leading him to the conclusion that his present avowal of fear fits into a pattern that causes his problems. Yet rational endorsement of this fact alone does not suffice to account for the healing effect of the insight he experiences during the session. It does not explain why he experiences it as a breakthrough in his therapy that opens up possibilities for change (rather than, e.g., as a confirmation of his hopeless situation, his shameful existence, etc.). The acknowledgment of his avowed fear from the second-person perspective also requires attitudes of acceptance, trust, compassion, encouragement, and so on, that tell him that it is OK to feel this way, that he is able to face his fear and should no longer run away from it, that he can do something about it, and so on. Following the idea of ‘internalization,’ these self-directed attitudes will in turn have been nourished by similar attitudes of his therapist toward him in the course of therapy.8

On our account, then, our patient has to fully experience his fear of abandonment and commit himself, through first-personal, affective avowal, to the expressions and descriptions that most clearly resonate with it. But affective avowal alone is not sufficient to realize the therapeutic shift in his approach to his fear; self-regulation from an internalized second-person perspective is essential as a safeguard against pathologies of avowal. Our patient also needs to acknowledge the affective avowal of his fear by adopting a particular range of allocentric attitudes toward himself: attitudes with a generally benevolent character that open up the possibility of controlling this painful emotional state in the future and of integrating it into a more encompassing, differentiated, and realistic view of himself, others, and the world.9


Our aim in this article was to make sense of the idea that insight can have a healing effect in psychotherapy. The question was how to conceptualize self-knowledge of one’s mental states such that we can understand its potential as a therapeutic factor. We argued for a dual perspective account of therapeutic self-knowledge. Accordingly, insight into one’s own mental states requires that one transparently avows them from the first-person point of view, while also relating to these avowed states from a generally benevolent, second-person perspective on oneself.

A next step would be to compare and integrate our philosophical account with theories of self-knowledge developed in the psychotherapeutic tradition. A full comparison with the psychotherapeutic literature obviously lies beyond the scope of this article. As we briefly mentioned, many psychotherapeutic traditions in one way or other make reference to a process of ‘internalization’ of interpersonal dynamics in the course of ontogenetic development and in the therapeutic process. Notions like ‘self-soothing’ (e.g., Kohut, 1977; Winnicott, 1953), the ‘holding environment’ (Winnicott, 1965), and its relation to ‘internal [End Page 55] (self-)object relations’ (e.g., Clarkin, Yeomans, & Kerberg, 2006; Fairbairn, 1952) in the psychoanalytic tradition, ‘internal working models’ in attachment theory (e.g., Bowlby, 1988), ‘schema modes’ in schema focused therapy (e.g., Young, Klosko, & Weishaar, 2003), and ‘the observing self’ in acceptance and commitment therapy (e.g., Hayes, Luoma, Bond, Masuda, & Lillis, 2006), for example, all point toward the relevance for psychotherapy of the second-person in oneself, that is, the allocentric self-relation.

IAt the same time, all therapeutic traditions stress the importance of experiencing one’s thoughts and feelings from the first-person point of view. It is an explicit focus in emotion-focused and client-centered psychotherapies (e.g., Gendlin, 1996; Greenberg, 2002). From a psychoanalytic perspective, Eagle (2011) contrasts first-person self-knowledge and ‘emotional insight’ with third-person self-knowledge and ‘intellectual insight,’ arguing that the former is the motor of insight-oriented psychotherapy. This contrast to some extent parallels the distinction between attending to ‘hot’ versus ‘cold’ cognitions in cognitive therapy (e.g., Safran & Greenberg, 1982). In cognitive therapy, patient and therapist collaboratively try to focus on ‘hot,’ affectively laden cognitions that are vividly experienced from the first-person point of view, as opposed to ‘cold’ cognitions that invite therapeutically ineffective theorizing about oneself.

It would be interesting to see whether, on closer examination, these and other conceptions of self-understanding in psychotherapy imply adopting a dual first- and second-person perspective on one’s mental states, as we have argued here. To take just one example, consider the concept of ‘mindfulness,’ as used in dialectical behavioral therapy (e.g., Linehan, 1993), mindfulness-based cognitive therapy (e.g., Segal, Teasdale, & Williams, 2002), acceptance and commitment therapy (Hayes et al., 2006) and other so-called third-generation behavioral therapies (Öst, 2008). Bishop et al. (2004) propose a two-component analysis of the concept of mindfulness:

The first component involves the self-regulation of attention so that it is maintained on immediate experience, thereby allowing for increased recognition of mental events in the present moment.

The second component involves adopting a particular orientation toward one’s experiences in the present moment, an orientation that is characterized by curiosity, openness, and acceptance.

(p. 232)

Bishop et al.’s description of the second component shows resemblance with our understanding of the allocentric perspective. Mindfulness seems to involve a particular range of allocentric attitudes toward oneself, including inquisitiveness, acceptance, and openness, excluding any kind of critical attitude toward one’s experience. It is an interesting question for future research whether the first component of mindfulness, that is, the ‘recognition of mental events in the present moment,’ can (or should) be understood in terms of (affective) avowal of the relevant mental events, or whether it involves some other kind of first-person self-knowledge.

All psychotherapeutic accounts have developed their own specific approaches as to how to promote patients’ understanding of their own mental states. These approaches, in turn, imply certain conceptions about what it means to relate to one’s own mental states in a healthy, adaptive way that can facilitate therapeutic change. The philosophical account developed in this paper might help to elucidate to what extent these conceptions converge and where they come apart. More important, perhaps, our account might help to understand why these modes of self-understanding described in the psychotherapeutic literature are therapeutically helpful.

Derek Strijbos

Derek Strijbos is a Postdoctoral Research Fellow at the department of philosophy of the Radboud University and psychiatrist at Dimence Mental Health Organization in The Netherlands. He wrote his dissertation on the non-representational dimensions of folk psychology. His current research focuses on various themes on the intersection of philosophy and psychiatry.

Fleur Jongepier

Fleur Jongepier is a PhD student at the Radboud University Nijmegen, and is writing her dissertation on self-knowledge and first-person authority.


The authors wish to thank the members of the Dutch philosophy and psychiatry network, Quassim Cassam, and two anonymous referees for valuable discussion and comments. Fleur Jongepier’s work on this paper was supported by the Niels Stensen Foundation.


1. Actual deliberation is not always required for a person to be granted first-person authority. What is essential, however, is that there is ‘logical room’ for the question whether wan is to believe that P, that is, that one’s mental states are ‘answerable’ to such considerations when called for (p. 63). [End Page 56]

2. The technique of ‘focusing’ in client-centered experiential psychotherapy provides an adequate illustration. See, for example, Gendlin (1996).

3. An argument could be made that, due to mental holism, failure to appreciate the (diachronic) mental context in which a mental state arises has negative impact on one’s self-knowledge of that state per se. Considerations of mental holism seem to undermine a clear distinction between the security and adequacy of self-knowledge.

4. For example, Systems Training for Emotional Predictability and Problem Solving (STEPPS; Black, Blum, & St. John, 2009).

5. All evidence-based psychotherapies for borderline personality disorder stress this self-regulative dimension of relating to one’s dispositions to affectively or deliberatively avow: Cf. dialectical behavioral therapy (Linehan, 1993), transference focused psychotherapy (Clarkin et al., 2006), schema therapy (Arntz & Van Genderen, 2011), and mentalization based treatment (Fonagy & Bateman, 2012).

6. Notice that allocentric endorsement of certain propositions about one’s first-order states does not imply avowal of the ensuing second-order belief. As we use the term (following the literature), avowal is necessarily first personal and always has a self-ascriptive aspect to it (e.g., I believe that it is the case that P) that distinguishes it from mere endorsement of the content of a second-order state (e.g., “I tend to form paranoid beliefs about other people’s intentions when I do not get enough sleep”). We suggest that allocentric endorsement of a proposition about one’s own (first-order) states is analogous to, because it is developmentally dependent on, the endorsement by other people, from the second-person perspective, of such propositions about one’s (first-order) mental states.

7. Such second-person or ‘participatory’ approaches, have been much discussed in the debate on folk psychology and social cognition. They start from the idea that we fundamentally relate to others as agents with normative (and moral) statuses, rather than psychological subjects for theory-based empirical inquiry (e.g., Gallagher, 2001; Ratcliffe, 2007; Reddy, 2008; Hutto, 2008; Strijbos & De Bruin, 2012; Zawidzki, 2013).

8. It should be noted that allocentric self-regulation can also be pathological. Taking the idea of ‘internalization’ seriously, we should expect people to internalize both healthy and unhealthy interaction patterns that they engaged in with their caregivers. Patients who seek help from psychotherapists often suffer from dysfunctional patterns of allocentric self-relatedness.

9. It is an interesting question whether allocentric regulation can alter not only the role of first-personal avowal and the states avowed in one’s ‘mental economy,’ but also the content of these states. A similar question can be raised about the effect of allocentric regulation on the kind of avowal (e.g., deliberative vs. affective) of the relevant mental states. Unfortunately, we do not have the space to address these questions here.


Armstrong, D. (1981) The nature of mind and other essays. Ithaca, NY: Cornell University Press.
Arntz, A., & Van Genderen, H. (2011). Schema therapy for borderline personality disorder. Oxford: Wiley-Blackwell.
Bishop, S. R., Lau, M., Shapiro, S., Carlson, L., Anderson, N. D., & Carmody, J. (2004). Mindfulness: A proposed operational definition. Clinical Psychology: Science and Practice, 11, 3:230–241.
Bilgrami, A. (2006). Self-knowledge and resentment. Cambridge, MA: Harvard University Press.
Black, D. W., Blum, N., & St. John, D. (2009). STEPPS: A practical, evidence-based way to treat borderline personality disorder. Current Psychiatry, 8, 31–37.
Bowlby. J. (1988). A secure base: Clinical applications of attachment theory. London: Routledge.
Carmen, T. (2003). First persons: On Richard Moran’s Authority and Estrangement. Inquiry, 46, 3, 395–408.
Carruthers, P. (2011). The opacity of mind: An integrative theory of self-knowledge. Oxford: Oxford University Press.
Clarkin, J. F., Yeomans, F. E., & Kerberg, O. F. (2006). Psychotherapy for borderline personality: Focusing on object relations. Arlington, VA: American Psychiatric Publishing.
Dennett, D. C. (1987). The intentional stance. Cambridge, MA: MIT Press.
Eagle, M. N. (2011). From classical to contemporary psychoanalysis: A critique and integration. New York: Routledge.
Edgly, R. (1969). Reason in theory and practice. London: Hutchinson.
Evans, G. (1982). The varieties of reference. Oxford: Oxford University Press.
Fairbairn, W. R. D. (1952). Psychoanalytic studies of the personality. London: Tavistock.
Fonagy, P., & Bateman, A. W. (2012). Handbook of mentalizing in mental health practice. Washington, DC: American Psychiatric Publishing.
Gallagher, S. (2001). The practice of mind: Theory, simulation, or interaction? Journal of Consciousness Studies, 8, 5–7, 83–107.
Gendlin, E. T. (1996). Focusing-oriented psychotherapy: A manual for the experiential method. New York: Guilford.
Gopnik, A. (1993). How we know our minds: The illusion of first-person knowledge of intentionality. Brain and Behavioral Sciences, 16, 1:1–14. [End Page 57]
Greenberg, L. S. (2002). Emotion-focused therapy: Coaching clients to work through their feelings. Washington, DC: American Psychological Association.
Hayes, S. C., Luoma, J. B., Bond, F. W., Masuda, A., & Lillis, J. (2006). Acceptance and commitment therapy: Model, processes and outcomes. Behaviour Research and Therapy, 44, 1:1–25.
Hutto, D. D. (2008). Folk psychological narratives: The sociocultural basis of understanding reasons. Cambridge, MA: MIT Press.
Kohut, H. (1977). The Restoration of the self. New York: International Universities Press.
Lear. J. (2004). Avowal and Unfreedom. Philosophy and Phenomenological Research, 69, 2, 448–454
Linehan, M. M. (1993). Cognitive-behavioral treatment of borderline personality disorder. New York: The Guilford Press.
McGeer, V. (1996). Is ‘self-knowledge’ an empirical problem? Renegotiating the space of philosophical explanation. Journal of Philosophy, 93, 10, 483–515.
McGeer, V. (2008). The moral development of first-person authority. European Journal of Philosophy, 16, 1, 81–108.
Moran, R. (2001). Authority and Estrangement: An essay on self-knowledge. Princeton: University of Princeton Press.
Öst, L. G. (2008). Efficacy of the third wave of behavioral therapies: A systematic review and meta-analysis. Behavior Research and Therapy, 46, 3, 296–321.
Padesky, C. (1993, September 24). Socratic questioning: Changing minds or guiding discovery. Keynote address, European Congress of Behavioural and Cognitive Psychotherapies, London.
Ratcliffe, M. (2007) Rethinking commonsense psychology: A critique of folk psychology, theory of mind and simulation. Basingstoke: Palgrave Macmillan.
Reddy, V. (2008). How infants know minds. Cambridge, MA: Harvard University Press.
Safran, J. D., & Greenberg, L. S. (1982). Eliciting «hot cognitions” in cognitive behaviour therapy: Rationale and procedural guidelines. Canadian Psychology/Psychologie Canadienne, 23, 2, 83–87.
Segal, Z., Teasdale, J., & Williams, M. (2002). Mindfulness-based cognitive therapy for depression. New York: Guilford Press.
Strijbos, D. W., & de Bruin, L. C. (2012). Making folk psychology explicit: The relevance of Robert Brandon’s philosophy for the debate on social cognition. Philosophia, 40, 1, 139–163.
Taylor, C. (1985). Human agency and language: Philosophical papers. Vol. 1. Cambridge: Cambridge University Press.
Vierkant, T. (2012). What metarepresentation is for. In: M. J. Beran, J. L. Brandl, J. Perner, & J. Proust (Eds.), Foundations of Metacognition, 279–288. Oxford: Oxford University Press.
Winnicott, D. W. (1953). Transitional objects and transitional phenomena: A study of the first not-me possession. International Journal of Psycho-analysis, 34, 89–97.
Winnicott D. W. (1965). The capacity to be alone. In M. M. R. Kahn (Ed.), The Maturational Processes and the Facilitating Environment, 29–36. New York: International University Press.
Young, J. E., Klosko, J. S., & Weishaar M. (2003). Schema therapy: A practitioner’s guide. New York: Guilford Publications.
Zawidzki, T. (2013). Mindshaping: The Linchpin of the Human Socio-Cognitive Syndrome. Cambridge, MA: MIT Press. [End Page 58]

Additional Information

Print ISSN
Launched on MUSE
Open Access
Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.